Search

The predominant antienvelope cell-mediated cytotoxicity in HIV-1-infected patients is a direct form of antibody-dependent cellular cytotoxicity (ADCC) in which circulating NK/K cells armed with cytophilic antibodies comprise a cytolytic effector cell complex capable of destroying HIV-1-expressing targets. This non-MHC-restricted form of virus-specific cytotoxicity is present in most infected patients, with maximum activity in early disease, gradually declining with disease progression. This endogenous cytotoxicity provides a focal point in the design of interventive strategies involving immune-based therapies. In the first such attempts, the lymphokine interleukin-2 has been employed in an effort to augment these potentially beneficial cytolytic reactivities. The focus of this article is to present the rationale, early clinical results, and future direction of such therapeutic approaches and, in doing so, to illustrate how careful basic research findings can be applied to the design of rational therapeutic strategies.

We have developed a series of murine monoclonal antibodies to a region of the 120 kD envelope glycoprotein (gp120) of human immunodeficiency virus type 1 (HIV-1). This region has previously been implicated as a site for virus neutralization by antisera raised to recombinant proteins and by antibodies made to full-length gp120 purified from virus. The antigen employed was a synthetic peptide containing 15 amino acids, representing amino acid residues 308-322, RIQRGPGRAFVTIGK, of env gp120 (HTLV-IIIB isolate). Five of the monoclonal antibodies raised to this antigen have reactivity with gp120 from divergent strains of HIV-1 in Western blot assays. The two of these five which were tested with live cells infected with the divergent HIV-1 isolates IIIB, MN, and RF were specifically reactive by fluorescence analyses with cells infected with the MN and IIIB isolates. Four of the five monoclonal antibodies blocked the fusion of IIIB-infected cells with uninfected MOLT-4 target cells. The monoclonal antibody most reactive with MN-infected cells by fluorescence, #5025A, blocked the fusion of MN-infected cells with uninfected MOLT-4 cells. Four of the five monoclonal antibodies neutralized the IIIB isolate of HIV-1 in vitro, but none neutralized the MN or RF isolates at the levels of antibody tested (less than or equal to 50 micrograms/ml). Taken together these data indicate that monoclonal antibodies to the immunodominant neutralizing domain of HIV-1 gp120 display different levels of group reactivity depending on the assay system being examined.

The principal neutralizing determinant (PND) of human immunodeficiency virus HIV-1 is part of a disulfide bridged loop in the third variable region of the external envelope protein, gp120. Analysis of the amino acid sequences of this domain from 245 different HIV-1 isolates revealed that the PND is less variable than thought originally. Conservation to better than 80 percent of the amino acids in 9 out of 14 positions in the central portion of the PND and the occurrence of particular oligopeptide sequences in a majority of the isolates suggest that there are constraints on PND variability. One constraining influence may be the structural motif (beta strand--type II beta turn--beta strand--alpha helix) predicted for the consensus PND sequence by a neural network approach. Isolates with a PND similar to the commonly investigated human T cell lymphoma virus IIIB (HTLV-IIIB) and LAV-1 (BRU) strains were rare, and only 14 percent of sera from 86 randomly selected HIV-1 seropositive donors contained antibodies that recognized the PND of these virus isolates. In contrast, over 65 percent of these sera reacted with peptides containing more common PND sequences. These results suggest that HIV vaccine immunogens chosen because of their similarity to the consensus PND sequence and structure are likely to induce antibodies that neutralize a majority of HIV-1 isolates.

A retrospective study of 153 hemophiliacs infected with human immunodeficiency virus-1 (HIV-1) was performed to determine the clinical and immunological consequences of HIV-1 infection and the markers and cofactors associated with these changes. Nearly 80% of HIV-1-infected hemophiliacs have developed a significant reduction in their CD-4+ counts (less than 400 CD-4+ cells/mm3) with 40% having less than 200 CD-4+ cells/mm3 by the end of 1987. The rate of CD-4+ cell count decline was slightly greater in patients who have already developed the acquired immunodeficiency syndrome (AIDS) compared to those who have not (50 vs. 31 cells/mm3/6 months). Thrombocytopenia and older age were associated with a more rapid CD-4+ count deterioration, but the quantity of clotting factor utilized did not affect immunologic progression. In patients with less than 200 CD-4+ cells/mm3, the incidence of AIDS was significantly higher in adults (greater than 21 years old) compared to children/adolescents. Cytomegalovirus (CMV) seroprevalence increased with age but did not correlate with the amount of concentrated clotting factor used. Although there was no relationship between CMV status and progression to AIDS, CMV-seropositive patients were older and had a lower CD-4+ count. Thus the majority of HIV-1-infected hemophiliacs are developing progressive immune dysfunction measured by CD-4+ count decline. This drop in CD-4+ count significantly correlates with a risk for the development of AIDS in adults but not in children (less than 21 years old).

Among the varied cytotoxic immune reactivities elicited as a result of HIV-1 infection are two forms of non-MHC restricted cytotoxicity--namely, indirect and direct ADCC. Since these reactivities are directed at both HIV-1 infected as well as gp120 coated targets, there is a potential for anti-HIV-1 ADCC to play both a beneficial as well as a pathogenic role in the natural history of HIV-1 disease. Resolution of these issues will be of great importance to the development of future preventive and interventive therapeutic strategies for AIDS. The direct and indirect forms of ADCC described herein are, most probably, not unique to HIV-1. In theory, any viral disease, retroviral or otherwise, in which high titers of anti-envelope antibodies persist in an environment rich in Fc-receptor bearing effector NK/K cells would be likely to have some component of direct ADCC as part of the host anti-viral response. With this in mind it is imperative that those researchers involved in characterizing cellular anti-viral cytotoxicities do not mistake direct ADCC for another form of CTL activity. These two highly potent reactivities operate independently and are subject to different control mechanisms, both positive and negative, anti-viral ADCC has too long been regarded as a strictly in vitro phenomenon with no in vivo counterpart. Our studies demonstrating direct forms of ADCC in infected patients will hopefully have some impact in forcing a careful re-evaluation of this extremely important issue.

Administration of high-titered goat anti-FLV gp71 IgG to AKR mice during a narrow neonatal therapy "window" suppresses the early development of MuLV infectious cell centers (ICC) in spleen, thymus, and bone marrow. By 4-5 weeks of age ICC appear in spleen and marrow cell populations, rapidly increasing to plateau levels within 2 weeks in the absence of viremia. The thymus of treated animals remains devoid of detectable ICC throughout the 4-month period of testing. This pattern of ICC suppression occurs independently of virus neutralization as shown by the inability of F(ab')2 preparations, which retained neutralizing activity, to prevent early establishment of ICC. Immune IgG significantly decreases, but does not eliminate gp71 expression in all tissues tested. In control animals, ICC reside within a minor subpopulation of cortical, thymic T cells, whereas peripheral (i.e., splenic) ICC are totally devoid of conventional T cell, B cell, and macrophage phenotypic markers. Although thymocytes appear to be a major target of this antibody therapy, T-cell reactivity is not compromised.

The antibody-dependent lytic activity of Corynebacterium parvum-induced peritoneal exudate cells was examined in vitro by utilizing AD755a tumor targets and a homologous anti-AD755a hyperimmune serum. Maximum antibody-dependent cell-mediated cytolysis (ADCC) of tumor targets was achieved within 4 hours of incubation. ADCC activity was found primarily in the plastic nonadherent cell population and was greatly enriched following removal of phagocytic cells by carbonyl iron. Phenotypically, the cells active in short-term ADCC were Qa-5+, ASGM-1+, Thy 1.2+, and NK 1.1+ and were unaffected by treatment with Lyt 1.2, Lyt 2.2, MAC-1, or I-Ab antibodies plus complement. Cells active in antibody-independent lysis of AD755a targets were phenotypically identical to antibody-dependent effectors. Although indicative of a natural killer (NK) cell phenotype, C. parvum-induced effectors differed from "spontaneous" splenic NK cells in their relative sensitivity to anti-Thy 1.2 as well as to anti-NK 1.1 treatment. Unlike the IgG2a-dependent lysis of AD755a-derived cells by inflammatory macrophages, all IgG isotypes of antiAD755a serum were equally effective in ADCC mediated by C. parvum NK cells. Finally, treatment of C. parvum-inoculated animals with anti-ASGM-1 serum eliminated in vitro NK activity and abrogated the in vivo therapeutic effects of hyperimmune serum. These findings, together with other correlations detailed herein, strongly suggested that C. parvum-activated NK cells appeared to represent a unique subset of NK cells that can serve as potent effectors in the antibody-dependent killing of AD755a tumor cells.

We have studied the biologic and physical properties of a monoclonal antibody that binds to gp120, the exterior envelope glycoprotein of the human immunodeficiency virus (HIV) strain HTLV-IIIB. Designated 9284, the antibody possesses viral neutralizing activity and inhibits syncytium formation by infected cells. The antibody recognized a region of the polypeptide backbone previously described as an important neutralizing epitope. This region lies 307-330 residues from amino terminus of the glycoprotein. We have compared the biologic and physical properties of this antibody to those of the recently described 0.5 beta monoclonal antibody to gp120. The 0.5 beta antibody was biologically more potent and bound an epitope slightly downstream to that of the 9284 antibody. The antibodies did not differ significantly in their affinity for gp120. In competition studies, the 0.5 beta antibody was displaced by the 9284 antibody, but the binding of the latter was unaffected by 0.5 beta.

Ecotropic virus-producing cells in the bone marrow of the leukemia-prone AKR strain of mice were significantly enriched by fractionation on discontinuous density gradients of Percoll and were found in a low-density population of cells comprised predominantly of medium to large blast cells. The high ecotropic virus-producing low-density bone marrow cell population was also found to be significantly enriched in pluripotent stem cells, prothymocytes, and stromal elements. During the period of time defined by a window for successful leukemosuppressive immunotherapy of AKR mice, virus-producing cells were exclusively detected in this fraction of bone marrow cells, implicating the functional classes of cells coenriched in this fraction as both potential targets of anti-viral immunotherapy and responsible for the seeding of the spleen and thymus with infectious ecotropic virus.

Passive anti-viral immunotherapy greatly suppresses the incidence of spontaneous leukemia in AKR mice, rendering the thymus of successfully treated animals devoid of infectious ecotropic retrovirus. Reconstitution assays have determined that the thymic and splenic homing cells of the AKR bone marrow become ecotropic virus producers subsequent to their seeding of these hematopoietic organs and that in vitro depletion of gp71 expressing bone marrow cells reduces stem cell numbers without affecting prothymocyte content. In the thymus, a population of radioresistant cells, which phenotypically resemble cortical thymocytes, but are unique in their expression of high levels of H-2Kk antigen, have been found to produce high levels of both ecotropic and MCF virus and have been implicated as a putative therapeutic target cell population of anti-viral treatment. In addition, the failure of treated animals to reconstitute following lethal irradiation suggests that an immunotherapy-induced alteration occurs in the bone marrow of AKR mice.

In a 6-week clinical trial 4 dose regimens of 3'-azido-3'-deoxythymidine (AZT), a thymidine analogue with potent anti-viral activity against HTLV-III in vitro, were examined in 19 patients with the acquired immunodeficiency syndrome (AIDS) or AIDS-related complex (ARC). AZT was given intravenously for 2 weeks, then orally for 4 weeks at twice the intravenous dose. AZT was well absorbed from the gut and crossed the blood-brain barrier. Therapeutic levels were maintained with 5 mg given intravenously or 10 mg given orally every 4 h. Treatment was not limited by side-effects, the commonest of which were headaches and depression of white-cell counts. 15 of the 19 patients had increases in their numbers of circulating helper-inducer T lymphocytes (p less than 0.001) during therapy, 6 who were anergic at entry showed positive delayed type hypersensitivity skin test reactions during treatment, 2 had clearance of chronic fungal nailbed infections without specific anti-fungal therapy, 6 had other evidence of clinical improvement, and the group as a whole had a weight gain of 2.2 kg. Also, with the highest dose regimen cultures of peripheral blood mononuclear cells for HTLV III became negative.

Tetanus toxoid (TT) reactive CD4+ cells were infected with HTLV-IIIB and exposed to TT at various times throughout a 7-day interval. Acute infection per se failed to produce overt cytopathology. However, exposure of infected cells to TT resulted in a rapid loss of cell viability, an increase in viral p24 expression, and a decline in T-cell blastogenesis. To determine whether HIV infection of antigen presenting cells (APC) could impact on T-cell activation, virus infected APC were utilized to present TT to responsive CD4+ cells. Use of infected APC produced effects similar to antigen stimulation of infected T-cells. These results suggest that the conditions of antigen presentation during T-cell activation may provide an excellent opportunity for virus transmission which may produce maximal immune dysfunction. However, preincubating antigen specific T-cells with the virostatic agent 3'-azido-3'-deoxythymidine (AZT) could prevent most of these effects.

Many murine tumor models associated with murine leukemia virus(es) (MuLV) have been successfully treated by passive administration of antiviral antibodies. There is a large body of virus-negative tumors, however, which are lowly antigenic and thus refractory to such approaches. Therefore, an investigation was done for determination of whether such tumors could be rendered susceptible to passive serum therapy by introduction of MuLV antigens onto the tumor cell surface. For this purpose a 3-methylcholanthrene-induced fibro-sarcoma from inbred C57BL/6J mice was chosen. Following infection of the tumor in vitro with Friend murine leukemia virus (F-MuLV), the tumor was found to be susceptible to treatment with a high-titered heterologous anti-F-MuLV gp71 antiserum. The specificity of the treatment was determined by conduction of the therapy on the uninfected parental tumor, in which case there was no effect. In addition, therapy could be initiated at time points when demonstrable tumors were present and successfully treated animals were resistant to rechallenge with the infected tumor. Thus conversion of a lowly antigenic virus-negative tumor to an MuLV-positive antigenic tumor rendered such growths susceptible to immunologic management.

Cytophilic antibodies which mediate antibody dependent cellular cytotoxicity (ADCC) against envelope antigens of human immunodeficiency virus (HIV) can be found in seropositive individuals. In these experiments, sera from a wide spectrum of HIV infected patients ranging from asymptomatic to overt acquired immunodeficiency syndrome (AIDS) were shown to contain high titers of antibodies that mediate ADCC. Not only did patient antibodies bind to surface expressed viral antigens and mediate ADCC against cells chronically infected with human T-lymphotropic virus type IIIB (HTLV-IIIB), but also against cells infected with the divergent HTLV-IIIRF2 and HTLV-IIIMN viral isolates. Similar results were obtained with target cells bearing purified GP 120 from HTLV-IIIB and HTLV-IIIRF2, indicating that a major portion of the activity was mediated by anti-GP 120 antibodies. Consistent with this was the ability to absorb most of the group-specific ADCC activity from the serum of an HIV infected individual using affinity columns bearing purified HTLV-IIIB GP 120. The finding that human antibodies reactive against the HIV envelope glycoprotein mediate ADCC against cells chronically infected with divergent strains of HIV will have important implications in designing rational approaches to passive and active immunotherapy.

Currently available anti-human immunodeficiency virus type 1 (HIV-1) agents such as azidothymidine can prevent de novo virus infection in vitro but lack significant activity against chronically infected cells. Our laboratory has recently described glycoprotein (gp)120-specific cell mediated cytotoxicity (CMC) present in HIV-1-seropositive individuals that is capable of destroying virally infected cells. As a means of potentially eliminating persistent reservoirs of HIV-1, we examined the ability of various cytokines to augment preexisting gp120-specific CMC activity of peripheral blood mononuclear cells obtained from early disease patients. We found that interferon-gamma alone had no effect on gp120 cellular reactivity; however, the combination of interferon-gamma plus IL-2 produced enhancement beyond that of IL-2 alone.

Although cortical thymocytes were found to be the predominant ecotropic and MCF-virus producers in the thymus of leukemia prone AKR mice the initial ecotropic retrovirus producing cells have been detected among a low density subpopulation of thymocytes including both PNA(+) and PNA(-) cells. Leukemosuppressive anti-viral treatment of these animals results in several important changes in the AKR thymus, including the elimination of virus producing cells, the induction of cellular phenotypic alterations, a decreased ability to bind ecotropic and MCF virus, a resistance to challenge with leukemogenic exogenous retroviruses, and the apparent elimination of a population of virus producing radioresistant cells. In addition, complement-mediated depletion with anti-viral IgG resulted in the complete elimination of proliferating thymocytes. These results suggest that passive anti-viral immunotherapy effectively eliminates a population of thymocytes which serve as the neonatal source of endogenous retroviruses and interferes with the earliest stages of leukemogenesis in AKR mice.

In efforts to identify B cell and T cell epitopes of HIV-1 structural components, serum as well as lymphocytes from HIV-1-seropositive individuals were reacted with several recombinant and native peptides representing defined viral gag and env determinants. Several areas of discordance between humoral and cellular reactivity were identified. Specifically, the principal neutralizing site within HIV-1, the major envelope glycoprotein gp120, failed to elicit detectable cellular reactivities. The carboxyl portion of gp120 and the transmembrane gp41 region were uniformly recognized by patient antibodies but did not produce significant lymphocyte blastogenesis. However, the amino half of gp120 elicited cellular responses in a majority of the immunocompetent individuals tested, despite its extremely low reactivity with patient sera. Last, the major HIV-1 structure component p24 was found to be the most consistent T cell activation antigen among the panel tested.

Antibody against viral gp71 is effective therapeutically for high leukemic AKR mice if injected immediately after birth. No corresponding effect could be observed after inoculation later in life when the endogenous virus burden is already high. However, if antibody treatment was supplemented by the injection of p15(E) antibody, a therapeutic effect was observed even in older mice first treated at an age of 21/2 months. Those mice produced antibodies against viral surface proteins and appeared to be able to survive longer than control mice. Thus p15(E) antibody might be able to overcome retroviral associated immuno-deficiency. This therapy may have implications for the treatment of the apparently retroviral induced acquired immunodeficiency syndrome (AIDS) of man.

Attempts were made to augment the antibody-dependent killing of the ascitic AD755a tumor in vivo to protect C57BL/6J mice against the outgrowth of larger tumor burdens. The lethal dose for this tumor is less than 100 cells, and antibodies contained in a hyperimmune antitumor serum (HIS) were found to suppress the outgrowth of a maximum of about 5 X 10(5) cells. Thioglycollate injected ip increased the number of peritoneal macrophages, potential effectors for antibody-dependent cell-mediated cytotoxicity (ADCC), by tenfold to fortyfold and raised the maximum treatable tumor challenge (MTTC) to about 4 X 10(6) cells. By comparison, ip injection of Corynebacterium parvum increased the total peritoneal cell population by only twofold but raised the MTTC to about 20 X 10(6) cells. Neither agent alone had an effect on long-term survival, even at very low tumor inocula (1 X 10(3) cells). The protective HIS is known to contain tumor-binding antibodies in each of the IgG1, IgG2A, and IgG2B isotype fractions. Although the IgG2A fraction is far superior in vivo in the suppression of tumor outgrowth, the IgG2A fraction was also found to be most effective in combination with thioglycollate treatment in agreement with the observed preference of thioglycollate-elicited macrophages for this isotype in in vitro killing assays. In contrast following C. parvum treatment, all three isoptype fractions were equally suppressive to tumor outgrowth. A second major change following C. parvum treatment was that tumor cells precoated in vitro with antibodies were effectively eliminated in vivo. The same antibody-coated cells administered to thioglycollate-treated or unmanipulated animals were uniformly lethal even at much lower tumor doses. Taken together these results suggested a major qualitative change in the antibody-dependent tumor-killing process following C. parvum treatment. This change was most likely due to the C. parvum activation of highly lytic effector cells for ADCC, the identity of which was examined in an accompanying manuscript.

A rapid, simple, reproducible and semi-quantitative assay to measure neutralizing antibodies has been developed. It employs a unique cell line which is exquisitively sensitive to infection with all HIV isolates tested. The assay is amenable to microtiter formulation as well as analysis by automation.

To examine the variables associated with heterosexual transmission of human immunodeficiency virus (HIV), we studied 32 couples in our hemophilia center who had steady sexual relationships for periods more than 1 year. Of the 32 sexual partners of the hemophiliacs, five (15.6%) were HIV seropositive. All five hemophiliacs with HIV transmission to their sexual partners had measurable immunologic deficiencies, as shown by their lower median T-helper (CD-4+) lymphocyte count of 172 cells/mm3. The hemophiliacs without transmission had a slightly higher median CD-4+ count of 297 cells/mm3 (P = .26). To determine if factors other than the degree of immunologic deficiency in the hemophiliac might contribute to HIV transmission, 18 of the 32 couples were studied more intensively by confidential, coded questionnaires. Regular condom use was reported by nine couples (50%). Two of nine women (22%) without condom usage acquired HIV. One of nine women (11%) using condoms was seropositive; she also reported eight needlestick injuries while assisting her spouse with clotting factor treatments. Intravenous drug abuse was reported in two of the five couples with HIV transmission. Thus, hemophiliacs are at risk for transmitting HIV parenterally as well as venereally. Despite various risk behaviours associated with HIV transmission, the prevalence of infection in our cohort of hemophiliacs' sexual partners is low and within the range (6.8-22%) reported by others. This study underscores the need for comprehensive education and counseling in what previously appeared to be a homogeneous clinic population at risk for transmitting HIV to others.

Immunization with either an Escherichia coli recombinant segment of the human T-cell lymphotropic virus (HTLV-III/LAV) envelope protein (gp 120) or with deglycosylated gp 120 envelope protein produced antibodies that neutralize HTLV-III/LAV infection in vitro. Virus neutralization titers of these antisera were equivalent to those obtained with purified native gp120 as immunogen. This localizes at least one class of neutralizing epitopes to the carboxyl-terminal half of the molecule. In addition, native gp120 prevented HTLV-III/LAV--mediated cell fusion, whereas the recombinant gp120 fragment did not. This shows that although glycosylation is not required for induction of neutralizing antibodies, it may be important for interaction with CD4, the virus receptor. A segment of the HTLV-III/LAV envelope produced in E. coli may be an important ingredient of a vaccine for acquired immune deficiency syndrome.

A 17-month-old boy in whom immune-mediated thrombocytopenia (ITP) was the presenting manifestation of infection with human immunodeficiency virus (HIV) is being successfully managed with intermittent high-dose intravenous gamma globulin (IVIG) allowing maintenance of hemostatic platelet counts while avoiding the immunosuppression associated with other therapeutic modalities used to treat ITP. He continues to demonstrate marked responsiveness to IVIG, and has been maintained on weekly or bimonthly infusions for 12 months. The serendipitous documentation of HIV infection prior to IVIG therapy for immune-mediated thrombocytopenia in this child documents the importance of HIV testing prior to IVIG therapy to prevent erroneous assignment of IVIG as the vehicle responsible for transmission of HIV infection. This case history also documents the importance of HIV testing in the diagnostic evaluation of immune-mediated thrombocytopenias.

The mechanism by which IgG2A from a syngeneic antitumor hyperimmune serum mediates destruction of target cells in the presence of thioglycollate-elicited peritoneal macrophages was investigated by using an in vitro assay system. Labeled tumor cells were found to exhibit a biphasic pattern of binding to the effector cells; this binding pattern was dependent on the presence of specific antibody. The initial binding phase produced no apparent changes in the target cell population. Target cells coated with specific antibody exhibited a similar early binding phase, but excess free antibody was required for the subsequent binding phase and its associated release of radiolabel and cell destruction. Several features of this process observed distinguished it from more conventional forms of antibody-dependent cell-mediated cytoxicity. These included 1) the preference for antibodies of the IgG2A isotype, 2) the association of cell destruction with the release of nuclear but not cytoplasmic label, and 3) the requirement of excess free antibody for target cell killing.

Infection by an attenuated replication-competent murine retrovirus (Friend leukemia virus-FLV4), but not other non-transforming retroviruses, stimulated rejection of transplantable thymomas (RL-cell line) and subsequent tumor immunity in syngeneic mouse recipients. FLV-infected RL-cells (RL-FLV) were unaltered in their in vitro growth, and grew progressively to kill sublethally irradiated animals and nude mice. Primary RL-FLV rejection was due to induction of increased natural killer (NK)-cell activity limited to peritoneal sites of tumor inoculation with a minor cytolytic macrophage population. Syngeneic mutant beige (NK-deficient) mice similarly rejected RL-FLV cells with increased peritoneal NK-cell activity and acquired immunity to the parental RL-tumor. While RL-FLV stimulated far greater peritoneal NK activity than did other tested retrovirus-infected RL-cells, the inherent susceptibility of these cells to lysis by normal NK cells was not altered by virus. RL-FLV induced NK effectors showed an indiscriminate lysis pattern that was independent of target cell type and retrovirus expression.

The bone marrow of AKR mice is the richest source of infectious ecotropic cell centers (ICCs) during the neonatal period. The bone marrow ICCs reside in a low-density population expressing high levels of viral glycoprotein (gp71) and Class I histocompatibility antigens (H-2Kk). In addition, ICCs are enriched in the lymphoid band of Ficoll-Hypaque-fractionated bone marrow, the adherent population of nylon wool separated cells and among the low-density subpopulation of Percoll-fractionated marrow. The observed dichotomy between viral antigen expression and actual virus production suggests that actively cycling cells may be the primary virus producers in the AKR bone marrow. The phenotypic and physical data indicate that bone marrow stem cells and/or prothymocytes may be among the initial virus producing cells in the AKR bone marrow. Leukemosuppressive antiviral immunotherapy delays the appearance of ICCs in the bone marrow but does not exert any major long-term changes on the populations of cells present.

Tissue distribution of radiation leukemia virus (RadLV) was examined after its inoculation into normal C57BL/6J (B6) mice, B6 mice bearing a transplantable, non-virus-producing thymic lymphoma (RL12-NP), and B6 mice bearing a transplanted non-virus producing, Harvey murine sarcoma virus-transformed fibrosarcoma. Virus expression was determined by competition radioimmunoassay for murine leukemia virus (MuLV) p30 (predominant group-reactive antigen of MuLV) and for RadLV p12 (a highly type-specific MuLV polypeptide) and by membrane immunofluorescence for cell surface gp71 (predominant envelope glycoprotein of MuLV). Normal adult B6 mice were given three sequential iv injections of RadLV and were examined several times up to 200 days later for the appearance of neoplastic disease or expression of virion antigens. No clinical abnormalities were noted, and animals remained healthy for greater than 200 days. Significant levels of MuLV p30 and RadLV p12 were detected only in the thymuses. Organs and tumors from RL12-NP-inoculated animals contained low or nondetectable levels of virion antigens. Inoculation of mice with RL12-Rad, a cell line derived by in vitro infection of RL12-NP cells with RadLV, produced widespread, discrete metastatic tumors and infiltrated the lymphoid organs of B6 mice in a pattern identical to that observed after administration of RL12-NP cells. Lymphoid organs of RL12-Rad-inoculated animals expressed variable levels of virion antigens reflecting differences in the extent of tumor cell infiltration as opposed to virus spread from tumor to host cells. Administration of infectious RadLV systemically into RL12-NP tumor-bearing animals converted these tumors to viron antigen expressors with levels in superinfected tumors equivalent to those found in RL12-Rad-induced tumors. Infection was highly selective, and host tissues were minimally contaminated by the inoculated virus. Part of this selectivity was explained by the thymotropic property of RadLV. A rapidly dividing murine fibrosarcoma was not infected by RadLV, but this same non-virus-expressing tumor could be infected by common fibrotropic MuLV isolates.

Forty-one patients seropositive for human immunodeficiency virus type 1 (HIV-1) were assessed for cell-mediated cytotoxicity (CMC) against autologous target cells bearing the major envelope glycoprotein of HIV-1, gp120. Effector lymphocytes from over 85% of seropositive patients showed CMC specific for gp120-coated targets, whereas seronegative individuals had no detectable CMC. As a group, symptomless individuals had the highest levels of CMC; patients with AIDS-related complex and AIDS showed progressively diminished reactivity. The gp120-specific CMC was mediated by a population of non-T-cell effectors phenotypically resembling NK/K cells. Cytolysis was not restricted by major histocompatibility complex determinants, as shown by killing of heterologous gp120-adsorbed targets and of HIV-1-infected cell-lines. Gp120-specific CMC was highly augmented in the presence of interleukin 2, so it may be possible to develop therapeutic strategies aimed at destruction of virus-producing cell reservoirs in infected individuals through stimulation of HIV-specific host CMC.

The magnitude of immunologic defects observed in HIV-1-infected individuals before the development of overt AIDS is disproportionately high in comparison to the levels of infectious virus in these patients--suggesting that factors other than direct virus-induced cytopathology may be involved. With this in mind, we investigated the immunologic consequences of the interaction between purified HIV-1 gp120 and the CD4 molecules expressed by uncommitted as well as Ag-specific lymphocytes. HIV-1 gp120 exhibited a dose-dependent immunosuppressive effect on: 1) Ag-driven proliferation of cloned CD4+ lymphocytes, 2) OKT3-driven proliferation of cloned CD4+ lymphocytes, and 3) cytolytic activity of CD4+, EBV-specific CTL. Thus, HIV-1 gp120 can, in a manner similar to OKT4A antibodies, suppress T cell activation and the expression of cytolytic activities through its interaction with CD4. Additionally, activated CD4+ lymphoblasts can be rendered susceptible to immune cytolysis by virtue of their binding of purified gp120. This "targeting" of activated lymphoblasts can occur with levels of gp120 far below that which is needed to saturate all OKT4A-defined CD4 epitopes. Adsorbed gp120 could be demonstrated on the surface of these cells for up to 12 h, a sufficient time for interaction with host cytolytic elements. The data from these in vitro modeling experiments highlight one of many potential mechanisms of HIV-1 induced immunosuppression and lymphocyte destruction that can occur in the absence of infectious virus and that is based on the unique interaction between HIV-1 gp120 and its cellular receptor, CD4.

We have attempted to identify the antigen on L5178Y cells that is the target for peritoneal cytotoxic T-lymphocytes (CTL) generated in L5178Y-O (original) immunized and challenged DBA/2 mice during establishment of the L5178Y cell tumor dormant state. These CTL exhibited in vitro lytic activity against methylcholanthrene-induced L5178Y and P815Y cells as well as Gross virus-induced BALB/c cells but not against a variety of other H-2d-tumor cell lines. The pattern of susceptibility to cell-mediated immune cytolysis was identical to the pattern of AB-dependent complement-mediated cytolysis produced by a rabbit anti-L5178Y antiserum. Quantitative expression of surface H-2d determinants was not a limiting factor in tumor cell lysis by CTL. The degree of CTL lysis of the susceptible cell lines was directly related to the amount of tumor-associated antigen expressed on the cell surface. The pattern of in vitro susceptibility to CTL lysis correlated well with in vivo transplantation resistance. The L5178Y cell antigen target for both CTL and Ra-anti-L5178Y serum lysis is likely to be either an endogenous AKR ecotropic viral glycoprotein with a molecular weight of 71,000 or one of two cell surface determinants, at Mr 85,000 and 135,000. These higher-molecular-weight antigens are neither endogenous AKR ecotropic viral-induced nor murine leukemia virus group-specific precursor structural proteins but may be transformation antigens shared by the susceptible tumor cell lines.

Derivatives of the CEM T and WIL-2 B cell lines showed striking diversity in their responses to the HTLV-IIIB strain of the human immunodeficiency virus (HIV). Several stable phenotypic patterns could be defined, based on whether cells were permissive (P+, P-) for virus production, were sensitive or insensitive to cytopathic effects after infection by free virus (C+, C-), and whether they underwent fusion on contact with virus-infected cells (F+, F-). Although expression of CD4 was essential for infection by HTLV-IIIB, very low levels were sufficient for productive infection of WIL-2 derivatives. Conversely, some CEM T cell lines that expressed ample CD4, and which were able to bind virus gp120 and undergo fusion, did not support productive infection by free virus. One nonpermissive, CD4+ derivative of CEM could bind gp120 but failed to undergo fusion, suggesting an alteration in some membrane protein other than CD4 that is essential for virus entry and HIV-induced cell fusion. The AA2 derivative of the WIL-2 cell line is also described, which is remarkably permissive for HIV replication and exquisitely sensitive to virus cytopathic effect. The panel of related cell lines with different host-virus phenotypes could be useful for more precisely defining steps in the infectious cycle of HIV, and for identifying host cell genes and gene products that determine the outcome of HIV infection.

Reverse transcriptase was purified from human immunodeficiency virus (HIV). It utilized the artificial primer-template poly(rA)-oligo(dT)12-18 more efficiently than activated calf thymus DNA, poly(rI)-oligo(dC)12-18, poly(rC)-oligo(dG)12-18, or poly(rCm)-oligo(dG)12-18. Maximum activity was observed at pH 7.0 to 7.6 in the presence of 5 mM MgCl2 and 100 mM KCl. 3'-Azido-3'-deoxythymidine triphosphate competed with dTTP for binding to HIV reverse transcriptase. Different kinetic constants were obtained with different primer-templates. Km and Ki values of 2.8 and 0.04 microM, respectively, were obtained with poly(rA)-oligo(dT)12-18. The corresponding values were 1.2 and 0.3 microM, respectively, with activated calf thymus DNA and 0.3 and 0.01 microM, respectively, with extracted virus and native template. Inhibition of the host cell DNA polymerases alpha and beta was considerably weaker. The Km and Ki values obtained with activated calf thymus DNA as the primer-template were 2.4 and 230 microM, respectively, for DNA polymerase alpha and 6.0 and 73 microM, respectively, for DNA polymerase beta. 3'-Azido-3'-deoxythymidine triphosphate could also serve as an alternate substrate for HIV reverse transcriptase. The resulting incorporation of 3'-azido-3'-deoxythymidine triphosphate into poly(rA)-oligo(dT)12-18 caused chain termination and premature deceleration of the reaction. The terminated primer could not be elongated when incubated with dTTP and HIV reverse transcriptase.

A 3-methylcholanthrene [(MCA) CAS: 56-49-5]-induced fibrosarcoma cell line and its Friend murine leukemia virus-infected counterpart were assessed for their susceptibility to lysis by so-called "natural" effector cells in a series of 51Cr release assays. Detailed functional and phenotypic analysis of lytic effector cell populations from normal C57BL/6 mouse spleens revealed an identity most closely associated with natural cytotoxic cells. Neither tumor cell line was found to be sensitive to natural killer-mediated lysis. Additionally, virus infection of the MCA-induced fibrosarcoma cell line did not affect susceptibility to natural cell-mediated cytotoxicity.